pyxDamerauLevenshtein

Install the package using pip. Note that this project requires Python 3.9+, a C compiler (such as GCC), and Cython to build the high-performance Cython implementation.

pip install pyxDamerauLevenshtein

Use damerau_levenshtein_distance(seq1, seq2, max_distance=None) to compute the minimum number of operations (insertion, deletion, substitution, or transposition) required to transform one sequence into another.

• Supported types: str, list, tuple, range, and more.
• Performance optimization: If you provide a max_distance integer, the function will short-circuit and return max_distance + 1 if the true distance exceeds the threshold, avoiding unnecessary computation.

from pyxdameraulevenshtein import damerau_levenshtein_distance

# Basic usage
damerau_levenshtein_distance('smtih', 'smith')  # returns 1

# Using a threshold for short-circuiting
damerau_levenshtein_distance('saturday', 'sunday', max_distance=2)  # returns 3

# Works with non-string sequences
damerau_levenshtein_distance([1, 2, 3, 4, 5, 6], [7, 8, 9, 7, 10, 11, 4])  # returns 7

Use normalized_damerau_levenshtein_distance(seq1, seq2) to compute the ratio of the edit distance to the length of the longer sequence.

• Scale: 0.0 means sequences are identical; 1.0 means they have nothing in common.
• Note: This definition is the inverse of difflib.SequenceMatcher.ratio().

from pyxdameraulevenshtein import normalized_damerau_levenshtein_distance

normalized_damerau_levenshtein_distance('smtih', 'smith')  # returns 0.2

If you need to compare one sequence against multiple other sequences (e.g., a list of strings), use the _seqs variants of the API. These functions return a collection of distances corresponding to each element in the target sequence.

• damerau_levenshtein_distance_seqs(seq, seqs, max_distance=None): Returns raw distances.
• normalized_damerau_levenshtein_distance_seqs(seq, seqs): Returns normalized distance ratios.

from pyxdameraulevenshtein import damerau_levenshtein_distance_seqs, normalized_damerau_levenshtein_distance_seqs

array = ['test1', 'test12', 'test123']

# Raw distances
damerau_levenshtein_distance_seqs('test', array)  # returns [1, 2, 3]

# Normalized distances
normalized_damerau_levenshtein_distance_seqs('test', array)  # returns [0.2, 0.3333333333333333, 0.42857142857142855]